Tapas Are Free! Training-Free Adaptation of Programmatic Agents via LLM-Guided Program Synthesis in Dynamic Environments
Proceedings of the AAAI Conference on Artificial Intelligence · 14.03.2026 · 40 (35) · 29477-29485
DOI: 10.1609/aaai.v40i35.40189 ↗Kurz gefasst
Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 14.03.2026. Verfasst von Jinwei Hu und 3 weiteren Autor:innen.
Abstract (Originalsprache)
Autonomous agents in safety-critical applications must continuously adapt to dynamic conditions without compromising performance and reliability. This work introduces TAPA (Training-free Adaptation of Programmatic Agents), a novel framework that positions large language models (LLMs) as intelligent moderators of the symbolic action space. Unlike prior programmatic agents typically generate a monolithic policy program or rely on fixed symbolic action sets, TAPA synthesizes and adapts modular programs for individual high-level actions, referred to as logical primitives. By decoupling strategic intent from execution, TAPA enables meta-agents to operate over an abstract, interpretable action space while the LLM dynamically generates, composes, and refines symbolic programs tailored to each primitive. Extensive experiments across cybersecurity and swarm intelligence domains validate TAPA's effectiveness. In autonomous DDoS defense scenarios, TAPA achieves 77.7% network uptime while maintaining near-perfect detection accuracy in unknown dynamic environments. In swarm intelligence formation control under environmental and adversarial disturbances, TAPA consistently preserves consensus at runtime where baseline methods fail. This work promotes a paradigm shift for autonomous system design in evolving environments, from policy adaptation to dynamic action adaptation.
Zitieren
APA
Hu, J., Dong, Y., Sun, Y., & Huang, X. (2026). Tapas Are Free! Training-Free Adaptation of Programmatic Agents via LLM-Guided Program Synthesis in Dynamic Environments. Proceedings of the AAAI Conference on Artificial Intelligence, 40(35), 29477-29485. https://doi.org/10.1609/aaai.v40i35.40189
Vancouver
Hu J, Dong Y, Sun Y, Huang X. Tapas Are Free! Training-Free Adaptation of Programmatic Agents via LLM-Guided Program Synthesis in Dynamic Environments. Proceedings of the AAAI Conference on Artificial Intelligence. 2026;40(35):29477-29485. doi:10.1609/aaai.v40i35.40189
BibTeX
@article{Hu2026,
author = {Hu, Jinwei and Dong, Yi and Sun, Youcheng and Huang, Xiaowei},
title = {Tapas Are Free! Training-Free Adaptation of Programmatic Agents via LLM-Guided Program Synthesis in Dynamic Environments},
journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
year = {2026},
volume = {40},
number = {35},
pages = {29477--29485},
doi = {10.1609/aaai.v40i35.40189}
}
EUVIMED